论文部分内容阅读
目的对GM(1,1)模型、霍尔特双参数指数平滑预测模型和ARIMA模型在肺结核发病率预测中的效果进行比较。方法利用1980-2007年北京市肺结核的发病率分别建立GM(1,1)灰色预测模型、霍尔特双参数指数平滑预测模型和ARIMA模型,对建立的模型进行拟合。比较3个模型的拟合效果,同时利用ARIMA模型对2008年北京市的肺结核发病率进行预测。结果针对北京市肺结核发病率建立的GM(1,1)模型、霍尔特双参数指数平滑预测模型和ARIMA模型的平均误差率(MER)分别为15.11%、9.51%、9.52%,决定系数R2分别为0.935、0.964、0.969。结论 ARIMA模型对于隐含波动周期并且不稳定的循环型时间序列拟合效果优于GM(1,1)模型,对解决时间序列类型的肺结核发病率等资料有很好的实用价值。
Objective To compare the effects of GM (1,1) model, Holt’s two-parameter exponential smoothing model and ARIMA model in predicting the incidence of pulmonary tuberculosis. Methods The GM (1,1) gray prediction model, Holt’s two-parameter exponential smoothing prediction model and ARIMA model were established respectively according to the incidence of tuberculosis in Beijing from 1980 to 2007, and the model was fitted. The fitting results of three models were compared, and the incidence of tuberculosis in Beijing in 2008 was predicted by ARIMA model. Results According to the GM (1,1) model established in Beijing, the average error rate (MER) of Holt’s two-parameter exponential smoothing model and ARIMA model were 15.11%, 9.51% and 9.52% respectively. The coefficient of determination R2 Respectively, 0.935,0.964,0.969. Conclusion The ARIMA model is better than the GM (1,1) model for implicit volatility cycles and unstable cyclic time series fitting, which is of great practical value for solving the data of time series type of tuberculosis incidence.